deep-research

深度调研方法论（8步法）：将模糊主题转化为高质量调研报告。触发词：/deep-research、深度调研、帮我调研、调研一下、对比分析注意：如果用户需要的是可视化图谱而非报告，请使用 research-to-diagram skill。

Quality

83%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Quality

Content

77%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A thorough, well-sequenced research methodology with strong actionability and clear validation checkpoints. Its main gaps are token padding (version history, anecdote, redundant checklists) and broken progressive disclosure — referenced template files are not actually present in the bundle.

Suggestions

Add the missing templates/ bundle files (intermediate-outputs.md, comparison-framework.md, fact-card.md, report-structure.md) so the signaled references resolve; this is the single biggest progressive-disclosure defect.

Remove or collapse the 版本历史 table and trim the '典型事故' worked example to lift conciseness without losing operational value.

De-duplicate the时效性 and推导防护 rules that appear both inline in the steps and again in the 质量检查清单, keeping each rule in one canonical location.

Dimension	Reasoning	Score
Conciseness	The body is information-dense and prescriptive rather than explaining concepts Claude already knows, but it could be tightened: the 10-row 版本历史, the lengthy '典型事故' anecdote, and repeated时效性/推导 rules across both steps and checklists add tokens without operational value.	2 / 3
Actionability	Concrete save targets per step (e.g. '02_事实卡片.md'), an exact packaging command ('tar -czvf ~/outcome.tar.gz -C ~/Downloads/research <topic>'), structured tables (资料分层 L1-L4, 时效敏感性), and a concrete Step 6.5 instruction to launch an independent Agent via the Task tool make the guidance specific and executable.	3 / 3
Workflow Clarity	The 8-step sequence is explicit with per-step '→ 保存' checkpoints, multiple BLOCKING validation gates (Step 0.5, Step 2, Step 6.5), and a clear feedback loop ('校验发现错误 → 回溯修正事实卡片/推导过程') plus extensive checklists, matching the score-3 anchor.	3 / 3
Progressive Disclosure	A reference table signals one-level-deep template files ('templates/intermediate-outputs.md', 'templates/fact-card.md', etc.), but the templates/ directory and all four referenced files are absent from the bundle, so the signaled navigation is broken and cannot earn a 3.	2 / 3
	Total	10 / 12 Passed

Description

90%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A well-constructed description with explicit trigger guidance and active disambiguation from a sibling skill. Its only weakness is specificity: it names a single outcome rather than enumerating multiple concrete actions.

Suggestions

Add 2-3 concrete actions to the description (e.g., '拆解子问题、分层核验事实、构建对比框架、产出可追溯报告') to lift specificity from 2 to 3.

Dimension	Reasoning	Score
Specificity	States one concrete outcome ('将模糊主题转化为高质量调研报告') and names the domain ('深度调研方法论（8步法）'), but does not list multiple specific concrete actions like the score-3 anchor requires.	2 / 3
Completeness	Clearly answers what ('将模糊主题转化为高质量调研报告') and when (explicit '触发词' trigger list), satisfying the explicit-trigger requirement that would otherwise cap at 2.	3 / 3
Trigger Term Quality	Explicit, natural trigger phrases users would actually say ('深度调研', '帮我调研', '调研一下', '对比分析') plus the slash form '/deep-research', giving good coverage.	3 / 3
Distinctiveness Conflict Risk	Has a distinct research-report niche and an explicit disambiguation note ('如果用户需要的是可视化图谱而非报告，请使用 research-to-diagram skill') that actively reduces conflict with a neighboring skill.	3 / 3
	Total	11 / 12 Passed

Validation

93%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: nexu-io/nexu
Path: apps/desktop/static/bundled-skills/deep-research/SKILL.md
Commit: dadfb1c

Reviewed: about 15 hours ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.